danish fracture database – dfdb -...

37
Danish Fracture Database – DFDB.DK Herlev Sygehus 04.03.2015 Kirill Gromov, MD, PhD, - e-mail: [email protected]

Upload: truongtruc

Post on 02-Feb-2018

224 views

Category:

Documents


2 download

TRANSCRIPT

Danish Fracture Database – DFDB.DK

Herlev Sygehus 04.03.2015

Kirill Gromov, MD, PhD, - e-mail: [email protected]

”Web-based prospective quality monitoring of fracture related surgery”

Evaluation: Rates of reoperation

Operative parameters

Purpose of DFDB

2

Why assessment of quality is relevant:

Life time prevalence of getting a fracture is 38.2 %

20,000 fracture related surgeries are estimated to be carried out annually in DK

No prospective data for assessing the quality of fracture related surgery in general has previously been collected

Assessment of Quality

3

Need for registry based research

• Observational data • Treatment guidelines • Implant monitoring

6 years ago: Paper and pen-based database tested at

Aabenraa Hospital

2 years ago: Web-based database initiated at Hvidovre and

Odense Hospitals, Dept. of Orthopaedics

Now: 21 departments in total

Administration and statistics

» Steering group with

annual meetings

Michael Brix Anders Troelsen

History of DFDB

5

21 departments n ≈ 35,000 operations

Kolding

Aabenraa

Herlev

Bispebjerg

Esbjerg

Holbæk

Viborg

Horsens

Aalborg & Farsø

Randers

Hvidovre

Odense

Rigshospitalet

Slagelse

Køge

Nykøbing F.

Hillerød

Århus

Vejle

Gentofte

Participating departments

6

Early data completeness and validity

COMPLETENESS Hvidovre Odense Both Departments

Primary surgeries 90% (145/161) 86%(102/118) 88% (247/279)

Reoperations 84% (27/32) 71% (22/31) 77% (49/63)

DATA VALIDITY App. 90-100 %

Gromov et al., DMJ, oktober 2013

Completeness and validity

7

> 8000 entries / 6 months

History of data entries

DFDB

8

• DFDB.dk online registration, developed by PROCORDO®

• 2 min per registration

• Filled out by the surgeon immediately after surgery

• Patient, trauma and surgery related factors are registered

• Reoperations are linked with primary operations

Structure

9

• Departments own their own data

• Department can use data for quality monitoring, research etc.

• Departments can veto use of data in projects by other departments

• Departments deliver data for annual report at the DOS congress

• All rules followed, all is reported to The Danish Data Protection Agency

• Alle departments have a person responsible for data who is also member of the steering group

Data

10

User featers

• Status messages • Surgical log

Kirill Gromov, læge, PhD, - e-mail: [email protected]

Status messages

• Landsdækkende database (ansøgning til SSI indsendt)

• Flere forskningsprojekter

• Implantat sporing

Fremtid

13

Forskningsprojekter

• 4 artikler i Pubmed • 1 accepteret til JBJS – surgical delay.

– Danish Fracture Database Collaboration ordning.

• Flere på vej ???

Surgical Delay 3517 primary surgeries for low energy, non-pathological AO31A and AO31B fractures on patients age > 50 were identified from DFDB. Primary Endpoint : 30-day and 90-day mortality (Civil Registrational system). Modifiable Risk Factors: Surgical delay = From diagnostics (x-ray) to onset of surgery. Educational level of surgeon AA = Attending or above BA = Below attending

Nyholm et al. JBJS. Accepted 15

Median age: 82.0 (range 51-107) AO31A (trochanteric) 1729 (49%) Female 2458 (70%)

16

Mortality

Surgical delay

Surgical Delay

Logistical regression* *Adjusted for age, gender, ASA score, fracture type, treatment, educationale level of surgeon/surgical delay

Surgical delay

30-day mortality risk (OR)

90-day mortality risk (OR)

≥12 vs <12h 1.45* 1.10

≥24 vs <24h 1.34* 1.23*

≥36 vs <36h 1.32 1.21*

≥48 vs <48h 1.56* 1.36

≥72 vs <72h 1.23 1.09

Surgical delay

30-day mortality risk (OR)

90-day mortality risk (OR)

<12 h 1 1

12-24 h 1.33 1.02

24-36 h 1.60* 1.22

36-48 h 1.37 1.08

48-72 h 2.61* 1.83*

≥72 h 1.67 1.18

Educatilnal level of surgeon

30-day mortality risk (OR)

90-day mortality risk (OR)

AA 1 1 BA 1.28* 1.26*

Conclusions

Short surgical delay (<12h) reduces risk of 30-day mortality.

The effect is less pronounced at 90-day mortality.

Junior surgeons should be supervised.

19

IMPLANTATSPORING Hvidovre Hospital

Intelligent implantat sporing &

klassifikation

Kirill Gromov, læge, PhD, - e-mail: [email protected]

Kirill Gromov, læge, PhD, - e-mail: [email protected]

Implantat sporing Behov

20

”Det anvendte implantat skal kunne identificeres entydigt og specifikt ved angivelse af implantatets

navn og model, størrelse af implantatet, fabrikantens navn, serienummer eller LOT

efterfulgt af batchkode”

1) Hvad har vi opereret patienten med? • Reoperationer • Recalls

2) Hvordan går det med det nye implantat?

• Kontinuerlig kvalitets monitorering

Kirill Gromov, læge, PhD, - e-mail: [email protected] 21

Implantat sporing idag

1) Hvad har vi opereret patienten med? • Reoperationer • Recalls

Kirill Gromov, læge, PhD, - e-mail: [email protected] 22

Implantat sporing idag

2) Hvordan går det med det nye implantat?

• Kontinuerlig kvalitets monitorering

• Nem tilgængelig og med landsdækkende potentiale

• Løbende monitorering med ”early warnings” • Intelligent implantat sporing med

klassifikation

23

Implantat sporing i fremtiden

Kirill Gromov, læge, PhD, - e-mail: [email protected]

Hvorfor klassificere?

Er det ikke nok at kende firmaet og lot nummeret nummeret?

24 Kirill Gromov, læge, PhD, - e-mail: [email protected]

1978

Ford Pinto – ”Barbecue that seats four”

25 Kirill Gromov, læge, PhD, - e-mail: [email protected]

Defekt produktion, design eller noget helt tredje?

26 Kirill Gromov, læge, PhD, - e-mail: [email protected]

27

2010

Defekt produktion, design eller noget helt tredje?

28 Kirill Gromov, læge, PhD, - e-mail: [email protected]

Indføring af cladestik

29 Kirill Gromov, læge, PhD, - e-mail: [email protected]

• Sammenligning på tværs af producenter

• Mulighed for at kigge på flere faktorer samtidigt

• Identificering af fejl i designet • Mulige farlige ”kombinationer”

30

Implantatklassifikation giver større viden

Kirill Gromov, læge, PhD, - e-mail: [email protected]

Cladestik ved implantatregistrering

• Klassificeres første gang af kirurgen • En implantat skal kun klassificeres en gang • Alle implantater klassificeres • Valideres af superbruger

31 Kirill Gromov, læge, PhD, - e-mail: [email protected]

Overlevelses kurver

32

Overlevelses kurver

33

Aktuel status på Marvsøm

• Hvad er evidensen for brug af forskellige implantater til proximal femur frakturer?

Begrænsninger:

– Engelsksproget – publiceret 1990 eller senere. – Trochantær eller collum fem (dvs. ikke pipkin eller distalt for troc.

minor) – Primære hoftefracture – Ikke patologisk – Angive failure of implant – Min.follow-up 1 år – Klart angive implantat

Søm

Implant N (studies)

N (patients) Prospective

N (patients) Retrospective

N (patients) total

Gamma3 12 759 259 1018

IMHS 7 210 352 562

InterTan 5 585 10 595

PFN 13 785 183 968

PFNA 20 887 686 1573 PTN 0 0 0 0

•HUSK AT REGISTRERE I DFDB